Reporting Signal Evaluation Framework

Classification Categories

Reverseau uses a structured classification system. When submitting a report, users select a caller type category that reflects their experience. The standard categories are:

• Scam — the caller attempted to obtain money, personal information, or access through deception
• Spam — unsolicited automated or mass-dialled calls with no legitimate purpose
• Telemarketer — commercial sales calls, whether solicited or unsolicited
• Robocall — automated pre-recorded messages without a live caller
• Safe — the caller was identified as a legitimate entity (business, government, known contact)

These categories are not authoritative legal or regulatory determinations. They reflect aggregated user-submitted assessments based on individual call experiences.

Category labels reflect user-selected reporting inputs and aggregated reporting patterns, not independent verification by Reverseau.

Consensus Determination

A phone number's displayed classification reflects the consensus of all reports received for that number. The process operates as follows:

1. Report collection — individual reports are received and processed through the reporting pipeline
2. Category weighting — each report contributes one equal classification input for its selected caller type
3. Majority classification — the most frequently reported category across submitted reports is surfaced as the primary displayed classification
4. Aggregated report rating — summary rating indicators reflect aggregated user-submitted assessments

Volume & Confidence

Classification stability generally increases with report volume. The dataset reflects the following confidence characteristics:

• Single report — represents one individual's experience; should be treated as a single data point rather than a definitive classification
• Multiple consistent reports — independent reports with the same classification produce stronger consensus signals
• High-volume numbers — numbers with many reports from diverse sources tend to produce more stable aggregated classifications

The platform does not assign numerical confidence scores. Instead, the report count displayed on each number page allows users to assess classification reliability contextually.

Recency Weighting

The displayed classification reflects the cumulative report history. Recent reporting activity is surfaced contextually through timestamps, activity indicators, and the recently updated feed. This allows users to distinguish between numbers with recent activity and numbers where reporting has been dormant.

Phone numbers can be reallocated by carriers over time. Historical reports may not reflect the current holder of a number — this limitation is documented in Data Limitations.

Mixed Classification Scenarios

Some numbers receive reports across multiple categories — for example, a number may have both "Scam" and "Safe" reports. In these cases:

• The majority category is displayed as the primary classification
• The distribution of all report categories is visible on the number's detail page
• Mixed classifications are common for numbers that have been reallocated, numbers used by large organisations with varied call purposes, or numbers where reporter experiences genuinely differ

Mixed classification does not indicate a system error — it reflects the inherent variability of community-reported data.

What This Framework Does Not Do

The reporting signal evaluation framework:

• Does not predict future caller behaviour
• Does not assign probability of fraud or scam activity
• Does not confirm or deny the identity of the caller
• Does not independently investigate or verify reported incidents
• Does not constitute legal, regulatory, or investigative determination

Classification reflects aggregated community reporting patterns within predefined display thresholds. It is an informational indicator designed for contextual awareness, not a definitive or investigative finding.

Interpretation Guidance

When reviewing classification data on a phone number page, consider the following:

• Review the total report count — higher volumes produce more stable classifications
• Consider report recency — older reports may not reflect current number usage
• Review the full distribution of reported categories, not just the primary classification
• Assess allocation metadata separately — carrier data provides context, not caller identity

Related Documentation

• Community Reporting & Processing Model — how reports enter the system
• Number Classification System — telecommunications numbering structure
• Data Limitations — interpretation boundaries